> But whose creative output consists solely of doing things a million people have already done? Is everyone else working on projects built exclusively out of lists of primes and rebalancing binary trees?
Yes. Most developers in the corporate world are building CRUD apps that are 90% boilerplate. Hopefully this helps explain the disconnect.
And even if GenAI progress stops here and it never gets better, it's incredibly useful. Why do people realize that it can't do EVERYTHING and then get stuck to the view that they can't use it for ANYTHING, confused as to why others are getting benefit from it?
Two things can be true (and I believe they are): the hype can be "staggeringly overblown" AND it can still be useful in many cases.